Moments after rebar placement 5

[howarts]

If a beam or column has certain moments and after you put rebars.. what would happen to the moments? For instance. A beam has certain moment at midpan, when you put the rebar to address the tension (bottom bars).. what would happen to the moments? I assume the beam would no longer deflect for the moments? But there should still be moment.. how much would be the retained moments or actual bending of the beam if there is sufficient tension bars?

 

[dhengr]

Wow... this is really scary. Talk about the blind leading the blind. And, a Masters Degree in Structural Engineering between them too, and they still couldn’t engineer their way out of a paper bag which was open at both ends. Give these two guys a whole bunch more moments between them, and they still won’t have a seconds worth of knowledge about reinforced concrete design.

 

[howarts]

Many structural designers don't know the meaning of neutral axis, strain and stress. If you will ask your structural designers in your area.. chances are you can meet many who don't know them or forgot the concepts (try to ask so you would believe me). I had to argue with a lot of them who are not aware of what they are. Remember most designers use software package so they don't try to understand the physics of it. This is reality.

 

[hokie66]

I think you are wrong, and sincerely hope so. Using software without understanding what it does, and not understanding the structures you are "designing" is dangerous, probably criminal, depending on where you are located in the world.

 

[howarts]

I know. That's why I haven't designed a structure yet. Would like to master the physics and principles. And watch others do it first and be apprehentice.

In our place. The designers are exempted from any liability if earthquake hits.. they can just blame it on the earthquake. And because of the many safety factor (1.2 D + 1.6 L), even bad structure can stand for service level gravity load. They
would only start falling if seismic wave hit them.

What region in the world has the designer sharing any liability if earthquake hits? There is no code that says your building must be magnitude 8.0 compliant (or other values), is there?

Most designers use software now.. and they hire operators who are ignorant in the concepts (I know this is wrong). So you can find in your location designers who don't understand the meaning of neutral axis of a beam or column. Try asking, and you shall know.

 

[mtu1972]

Howarts - You have repeatedly used the term Designer. Where I worked drafters progressed to be designers through their career. They often had the title "CAD Structural Designer" or other discipline's such as mechanical or electrical.

Yet many of them felt they were competent in doing fundamental designs. And I agree that they would not understand the fundamentals of concrete design. Any Civil Engineer who studied the structural option should understand how to design concrete beams.

gjc

 

[molibden]

I am liable - personally and my company also. For common loads or earthquake. Law says one must follow the code. In Italy recently they had some major earthquakes and some new buildings collapsed. I'm sure I heard they investigated and brought some people on court.
If a building fails in the seismic event, on court they would judge if I designed per code requirements. It is not my responsibility however to design for every seismic event and every accidental load that is not part of the code.
I work in EU so I'm under Eurocodes.

 

[dhengr]

The point is, that for the safety and protection of the general public, in any (and all) locations, people who don’t have sufficient technical training, engineering education and training, and who don’t have a good sound understanding of the fundamentals of the structures they are designing and how they act/work, should not be doing design of these structures or systems. And, just because they have access to some software which they probably don’t understand either, does not make this type of design work proper or ethical either. This thinking should not preclude a drafter or technician from learning some of the fundamentals and some basic design functions when under the direct supervision of a real, experienced engineer who does take responsibility for the adequacy of the design work. The OP’er. and his friend are acting very dangerously and irresponsibly, even if one of them asks a few questions on an engineering forum, since we can’t teach full Uni. courses in reinforced concrete design here.

 

[KootK]

One of the things that I've learned through my adventures here at Eng-Tips is that there really are parts of the world where the average structural designer skill level is both sorely lacking and utterly corrupted by an over-reliance on software. And the scariest part of it, in my opinion, is that it leaves many studious, conscientious, junior engineers without any access to quality mentoring. As I see it, there are only two realistic outcomes in the short term:

1) Those junior engineers get no quality mentoring and design terrible buildings. Who would stop them?

2) Those junior engineers get some mentoring from forums like this and design less terrible buildings.

The pragmatist in me is drawn to alternative #2. As such, I'll happily do my best to help out with some of the fundamentals here.

THE FALSE "RIGHTNESS" OF THE ELASTIC MOMENT DISTRIBUTION

As designers, we most often work with the results of linear elastic computer models built from prismatic elements. A moment distribution so obtained is predicated upon the following:

1) A material that behaves linearly elastically and;

2) Prismatic members.

In general neither of these things is strictly true with reinforced concrete. This is because reinforced members are generally cracked non-uniformly, reinforced non-uniformly, and subject to all manner of complex time dependent phenomena etc.

So why do we use the linearly elastic moment distribution? Here's a few reasons:

1) It's relatively easy to work out computationally. And we're lazy/efficient folk.

2) It's usually at least as valid as most any other moment distribution.

3) Because moments in the service range tend to be close to the linear elastic distributions, supplying reinforcing to match these distributions tends to minimize serviceability issues such as cracking.

4) Reinforced concrete is usually quite ductile but not infinitely so. Staying close to the linear elastic distribution tends to minimize the amount of ductility relied upon. Often codes will limit how far designers are allowed to deviate from linear elastic moment distributions. Deviations on the order of 15-30% are common depending on the jurisdiction.

The takeaway here is to recognize that the linear elastic moment distributions that our computers spit out at us are not gospel. Rather, they are but one of many possible moment distributions that can be used to successfully design continuous concrete members. We can use alternate moment distributions because modern reinforced concrete design is based on a model of plasticity rather than strict elasticity. The practical motivation for redistributing moments is to reduce them in areas where rebar congestion has become a problem. Most often, this is at top steel locations over supports.

STATIC MOMENT CONCEPT

But I'm still googling about static moments so hope Kootk can shed more light on it as IDS also requested.

Yeah, my bad. Google doesn't seem to turn up squat and it's all buried in stuff pertaining to moment of inertial etc. I must be using non-conventional terminology or something. I find it unthinkable that such a fundamental concept could be getting so little web air play. I'll just go ahead and outline the method myself. I've included some sketches below as well.

In a very general sense, we can do this as designers:

1) Pretend that a continuous member is simply supported and calculate the moments accordingly.

2) Add any value of end moment to either end of the beam and, in doing so, reduce mid-span values.

3) Super impose the two triangular moment diagrams from #2 end moments with the #1 moments and reinforce for that.

Naturally, this process is limited by the stuff that I described above: presence of adjacent continuous framing and practical/code limits on redistribution based on ductility. As a simplification of the method described above, the classic static moment method goes like this:

1) Assume simple support and calculate the moment diagram and mid-span moment Mo (the static moment).

2) Add in negative end moments as desired and reduce the mid-span moments accordingly.

3) To ensure satisfaction of equilibrium, ensure that (M_left + M_right)/2 + M_midspan = Mo. This can be done graphically by constructing the final moment diagram, connecting the end moments with a straight line, and ensuring that the vertical distance from the mid-span moment value to the line is still equal to the static moment (Mo).

I think that this is an important concept for you because your intuition seems to be telling you that moment can not just be assigned randomly. There needs to be some kind of "conservation of moment" like there is "conservation of energy" in thermodynamic systems. And you're right. The static moment (Mo) is the thing that must be maintained no matter what. It's your flexural design North Star so to speak. Many concrete codes still include a "Direct Design Method" for hand calculation. It usually starts with working out the static moment and then redistributing that moment along the span and, in the case of two way slabs, perpendicularly to the span. This ensures the satisfaction of equilibrium.

The takeaway here is to recognize that the normal process of design is to decide upon the moment field to be used and then to supply reinforcement accordingly. It is rarely the case that we select reinforcement and then attempt to work out a suitable moment distribution to suit that reinforcement. The main exception, as I mentioned previously, is when we're trying to shed some moment at a location where rebar congestion has become a problem. And even then, the process usually goes:

1) Decide upon moment.

2) Reinforce for moment.

3) Tweak moment a bit to reduce congestion.

RELATIONSHIP BETWEEN REINFORCING AND MOMENTS

If a beam or column has certain moments and after you put rebars.. what would happen to the moments? For instance. A beam has certain moment at midpan, when you put the rebar to address the tension (bottom bars).. what would happen to the moments? I assume the beam would no longer deflect for the moments? But there should still be moment.. how much would be the retained moments or actual bending of the beam if there is sufficient tension bars?

A more senior structural engineer told me when you put rebars in the beam, you straighten the beam and the moment/bending get less.. so I'm trying to find basis in his argument. But for beam with tension bars, the deflection is less. Isn't the moment same as deflection?

This concept is simply incorrect. Even though both are forms of strain, it is better to think in terms of curvature rather than deflection because the relation of interest is this:

M = k * EI (M = moment; k = curvature; EI = stiffness).

So, if you could somehow reduce the curvature without increasing EI then, yes, moment would be decreased by decreasing curvature (say.. 20%). However, how did you go about reducing deflection and curvature 20%? Probably by increasing EI 20%. So, when all is said and done, you've decreased curvature 20%, increased EI 20%, and changed the moment not at all. This is the case, at least, when all members of a continuous system have their EI values increased proportionally. If some members are stiffened while other are not, then moment will migrate to the stiffened members.

Let's take the example of a simple beam supported by two columns on the left and right. He stated that by putting more bars in the negative moments at support.. he can straighten the beam and control the moments at midspan and using less tension bars at midspan.

So long as a moment connection is detailed between the beam and column, this is true. The static moment requirement remains however. And it must be recognized that this is not merely a reduction of mid-span moment but a transfer of moment demand from the midspan to the ends of the beam. Moment is conserved as discussed above.

YOUR COLLEAGUES ODD BEAM DETAILS

I told him why the bars at midspan are not at bottom and mid height and he replied "the area covered by actual moment reaches into the mid height of the beam section at midspan so it's ok the rebars reaches up at mid height".

Your colleague may in fact be right. We'd need to review the framing plan in order to know for sure. Some reasons to reinforce the beam as detailed include:

1) If it's a beam more or less simple spanning between two supports, the top steel may just be nominal reinforcing for crack control.

2) In a 250 mm wide beam, having multiple layers of reinforcing with interior bar can cause issues with achieving proper concrete placement and consolidation.

3) So long as the vertical position of all reinforcing bars has been properly accounted for analytically, all bars below the neutral axis can indeed contribute to resisting flexural tension. While this may not be optimal flexurally, it is valid.

4) In seismic design, sometimes this arrangement is used intentionally, to advantage. Let's leave that discussion for another day however.



I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[rapt]

Kootk,

Definitely no stars for this one.

Based on your logic with a single span beam and end columns, I can decide on a moment at the ends of wl^2/8, add reinforcement for this continuous in the top (or even terminate some towards the centre if you want) and provide no bottom reinforcement!

What happens when the end columns do not provide sufficient stiffness to generate the moment at the ends?

There are severe limitations to the methodology you have suggested that require an experienced engineer to apply.

Leave the basic approach at design for elastic moment distribution and we will be a lot safer, and closer to the real thing as well once you get both negative and positive plastic hinges forming. If an engineer does not understand elastic moment distribution, he/she should get out of the industry.

 

[howarts]

The above is the layout. My friend told me there are many midspan bars in B7 (the horizontal beam at middle) because it has secondary beam crossing it (the vertical beams). But then in the support, he has only 4 negative moments bars (as I shown before) with the 2 in midheight and I'm afraid it may crack so I told the client months ago to reject the plan and find another structural engineer. So don't worry.. the plan was not the one constructed. I told the new structural engineer to put more bars at the support.. or over design it. So there is no problem.

My interest is physics. I studied structural engineering just to understand newtonian physics and to verify any design to make sure my existing house or office won't come falling down in the event of seismic activity. So don't worry. I won't build actual real structure.

Kookt. I'll try to analyze physics part of it. If you have actual reference about static moments. Do let us know so we can study it at more details. Thank you.

 

[howarts]

By the way (to add to my message above & clarify). Initially i mentioned my friend mentioning putting more support negative moment bars to straighten the midspan.. but when I checked the old plans. I was mistaken. It's the opposite. He was actually referring to putting more midspan bars to make the beam more straight so less bars at support. That is. More bars at midspan than necessary to support the secondary beam crossing it.

BAretired said this was not good because elastic moment can be more than the moment capacity at support and support secton can crack. No problem about it. But if you put more midspan bars than the actual load demands to really make it straight.. does this moment redistribution really make the support moment more straight? (Kootk, etc.?) I just want to know this and I'd never do this in practice because plastic moments can occur at support and agree with rapt it's not good to deviate from elastic design. This is why I reject my friend design because I don't deal with speculations or untraditional thoughts. Again I'm just interested in the principles and won't do it in practice.

 

[BAretired]

A flat plate might be more economical. It eliminates a lot of formwork. But that is a separate issue.

I think what KootK means by "static moment" is really simple span moment. Beams B-7 and B-8 are loaded with a concentrated load P at midspan and a uniform load W representing self weight. The simple span moment is PL/4 + WL/8. A moment distribution would provide the magnitude of the negative moments at supports A, B and C from which the midspan moments could be calculated. The fixed end moments are PL/8 + WL/12.

In doing a moment distribution, it is customary to consider the possibility of unbalanced live loads. It is not realistic to assume live load is evenly distributed over both spans simultaneously. To simplify hand calculations, some engineers use a technique called "two cycle moment distribution" which takes into account any imbalance they wish to consider.

BA

 

[BAretired]

By the way (to add to my message above & clarify). Initially i mentioned my friend mentioning putting more support negative moment bars to straighten the midspan.. but when I checked the old plans. I was mistaken. It's the opposite. He was actually referring to putting more midspan bars to make the beam more straight so less bars at support. That is. More bars at midspan than necessary to support the secondary beam crossing it.

BAretired said this was not good because elastic moment can be more than the moment capacity at support and support secton can crack. No problem about it. But if you put more midspan bars than the actual load demands to really make it straight.. does this moment redistribution really make the support moment more straight? (Kootk, etc.?) I just want to know this and I'd never do this in practice because plastic moments can occur at support and agree with rapt it's not good to deviate from elastic design. This is why I reject my friend design because I don't deal with speculations or untraditional thoughts. Again I'm just interested in the principles and won't do it in practice.

Adding bars at midspan does not make the beam straighter, thus justifying fewer bars at the supports. The beam will feel moments in accordance with elastic theory until the beam cracks. If the beam cracks at the supports because the reinforcement is inadequate, then an excess of steel in midspan may prevent failure but it is not considered good engineering practice.

One of my first projects involved the review of drawings of an old concrete building which had been standing for nearly fifty years. One feature which caught my attention was a continuous beam reinforced with only bottom steel, no top steel. The bottom steel was adequate to carry the simple span moment so, in effect, the beam was acting like a series of simple spans. Cracks formed over each support, forming a virtual hinge. It was not sound engineering practice, but the beam remained in service for a long time.

Building codes permit some deviation from elastic design but not that much.

BA

 

[howarts]

Adding bars at midspan does not make the beam straighter, thus justifying fewer bars at the supports. The beam will feel moments in accordance with elastic theory until the beam cracks. If the beam cracks at the supports because the reinforcement is inadequate, then an excess of steel in midspan may prevent failure but it is not considered good engineering practice.

Why would an excess steel in midspan prevent failure of the supports when the midspan bottom bars won't reach the support at all? Unless you mean it reaches the support (in your first review project of the continuous beam)? If it won't reach the support, then the support can fail and the beam would just fall down.

If Kootk and you are right. Then my friend is wrong to say that adding more bars at midspan can straighten the beam and requiring fewer bars at support. Thank you for this illumination!

 

[Okiryu]

This is an interesting topic. Appears to be more complicated than maybe at first look... but what Kootk drew in his big post is what I was taught in my first day in my reinforced concrete course in the university. That is nostalgic...

 

[howarts]

Thank you all for the pointers.